Full Idea
The fuller we make the description of a cause, the better our chances of demonstrating that it was sufficient (as described) to produce the effect, and the worse our chances of demonstrating that it was necessary. (For the effect, it is the opposite).
Gist of Idea
Full descriptions can demonstrate sufficiency of cause, but not necessity
Source
Donald Davidson (Causal Relations [1967], §3)
Book Reference
Davidson,Donald: 'Essays on Actions and Events' [OUP 1982], p.157
A Reaction
If the fullness of description is relevant, this suggests that Davidson is focusing on human explanations, rather than on the ontology of causation. If the cause IS necessary, why wouldn't a better description make that clearer?